Modelling and forecasting African Urban Population Patterns for vulnerability and health assessments (MAUPP)

Start-End 01/10/2014 -  30/09/2018
Programme STEREO 3
Contract SR/00/304
Objective The population of Africa is predicted to double over the next 40 years, driving exceptionally high urban expansion rates. Urbanization has profound social, environmental and epidemiological implications and makes spatial and quantitative estimations of urban change, population density and socio-economic characteristics valuable information for epidemiology and vulnerability assessment. The performance of urban expansion models largely depends on the quality and type of data available, which have so far been limited, and reduced the confidence and the applications of models for Africa. Satellite remote sensing offers an effective solution for mapping settlements and monitoring urbanization at different spatial and temporal scales and allows to link empirical observations with urban theory.  Moreover, remote sensing data have a great potential to map and predict intra-urban variations in population density because they provide information on the morphology of different residential patterns that can be linked to different population densities and socio-economic parameters.

We have highlighted two limitations in the existing population distribution datasets that are related to urban areas: (i) the lack of urban expansion forecasts and (ii) the homogenous distribution of people within cities. Both limitations can be significantly improved with advanced remote sensing data and techniques. The general objective of the MAUPP project is therefore to improve our spatial understanding and forecast of urbanization and urban population distribution in SSA through the use of remote sensing and spatial modelling.

The project addresses two specific objectives:

Produce an urban expansion model at moderate spatial resolution for African cities.

This objective will be achieved through
  1. using HRRS and VHRRS (both optical and radar) to delineate urban extent of a set of cities, compare the accuracy and limitations of the obtained features and optimize the method based on HRRS,
  2. using of the best methods identified using HRRS to generate a database of land cover change to urban over the last 30 years across a large number of cities in Africa,
  3. using this database to build urban expansion models, evaluate their forecasting accuracy, and apply them to forecast the future distribution of the major urban extents in SSA.

Understand and predict intra-urban variations in human population density in Africa.

This will be achieved through
  1. using HRRS and VHRRS to predict human population density within the urban extent of a set of cities, compare the accuracy and limitations of the RS products, and optimize the method based on HRRS,
  2. using VHRRS to analyse and to understand the drivers of changes in human population density within cities;
  3. using the best methods based on HRRS data to generate a database of human population density within urban extent across a large number of cities in Africa.

Method The MAUPP project will work at two different spatial resolutions: HR (~30 m) and VHR (< 5m). The Morphological Urban Area (MUA) and urban land uses will be delineated, mapped and predicted mainly using SAR and optical remote sensing as input data. In addition, spatially-detailed census data, historical maps and existing population distribution datasets will be used for modelling and predicting urban expansion and human population densities across Africa.

MUA delineation using HRRS data

At the continent scale, the MUA of 30 to 50 African cities will be delineated using combined HR archived SAR and optical data. Information fusion extracted from the two datasets will allow the exploitation of redundancy, in order to increase global information and complementarity and to improve certainty and precision. Multi-temporal processing will be conducted for the years: 1995, 2000, 2005, 2010 and 2015. Due to this large multi-temporal dataset of maximum 50 African cities, we propose to apply a rapid and entirely automatic approach to extract human settlement from the SAR and optical HRRS data. The processing method developed by Dell’acqua et al.  and which combines co-occurrence matrix and an updated histogram distance index analysis will be applied for mapping urban density using satellite SAR data. Information on urban areas extracted from SAR and optical data will be combined using SVM. At the end of the processing chain, a final delineation of the urban area will be extracted based on the maximum of membership values after the combination step. For each city, feature based change detection will be applied for processing and analyzing the multi-temporal urban extent.

Mapping land use using VHRRS data

At the city scale, the land use will be mapped within the MUA using a robust methodology based on combined very high resolution (VHR) SAR and optical data in an Object-Based Image Analysis (OBIA) approach. Pleiades and RADARSAT-2 data will be acquired for three selected African cities. Layover, shadowing and multiple-bounce effects are characteristic elements of VHR SAR scenes of urban landscapes. Several PolSAR decompositions will be applied  for extracting textural and morphological features from the scenes. These features will be directly used as input bands to the fusion classifier. Fused data will be segmented using the versatile and popular OBIA software eCognition with its Server licence allowing the processing of large datasets. The segmentation will be constrained by the relevant OSM layers (building, roads, …) .  For each region, eCognition will be used to extract features from optical (spectral, textural and morphological), VHR SAR features, and ancillary data (e.g. height). The relevant features will be chosen according to an ontologic approach .  For each city, African experts in remote sensing will define land cover interpretation keys and illustrate them with located ground pictures. These interpretation keys formalising the expert knowledge will then be used to select quantitative features. A common typology of residential land use will be proposed.  The residential land use types will be identified for each urban block by a quantitative discriminant analysis of landscape indices computed on the land cover map, complemented by other indicators such as the distance to the city centre or the environmental conditions (e.g. located close to a marsh or on steep slopes).

Modelling MUA changes

A generalizable urban expansion model will be developed at HR using the urban land cover change database generated for 30 to 50 cities and for the years 1995, 2000, 2005, 2010 and 2015. Boosted Regression Trees (BRT) models will be used to calculate a probability of rural to urban conversion for each non-urban 30m pixel based on a set a covariates such as the accessibility, the topography or the proportion of urban pixels in the neighbourhood. Cross-validation techniques will be used to calibrate and validate the model. Then, using urban growth rates predicted by the UN  and the probability raster, the evolution of the spatial pattern of the MUA will be predicted for each African city. Methods described here follow approaches already developed before by LUBIES but that suffered from limited urban change training datasets . Here, the large database created using HRRS data will allow the production of Africa-wide urban predictions more spatially detailed than what has been done before. Moreover, the MUA of the 3 cities will be mapped through time at VHR using a collection of historical maps since colonial time and the contemporary VHRRS land use – land cover map of the cities. African experts and relevant literature will complement this mapping effort; the aim being to document the history of the 3 cities and identify the geographic factors that contribute or limit their extension.  In addition, VHRRS data assembled for these 3 cities will allow to refine the covariates used in the HR model and test other geographic determinants of the MUA evolution such as roads, markets, central business district or non-aedificandi zones.

Mapping intra-urban population densities

The most recent and the most spatially detailed census data will be acquired for the 3 cities selected as case studies. Geo-statistical methods will be used to study the relationships between census-derived population densities and VHRRS-derived attributes of residential land use types. VHRRS-derived attributes will include
  1. landscape metrics computed using the land cover map,
  2. spectral, textural and morphological features derived from VHRRS data and
  3. geographical variables susceptible to explain the distribution of populations such as distance to the city center, distance to main roads, presence of colonial non-aedificandi areas, presence of military camp, etc.

Based on these relationships, population weights will be assigned to the different land use classes and populations will be redistributed within each administrative census unit according to these weights. A very spatially detailed population density map will then be created for the 3 cities.
Geo-statistical methods will then be used to explore relationships between population densities calculated using VHRRS data and HRRS-derived attributes such as
  1. landscape metrics derived from the HR urban maps,
  2. spectral and textural features derived from HRRS data and
  3. geographical layers available for the whole African continent such as road and railway networks.

The main HRRS-derived attributes related to population densities will be used to build a model to predict intra-urban variations in population density in any large African city. The model will be calibrated and validated using detailed population census data assembled by the WorldPop project.

Integration and dissemination

Throughout the project, efforts will be made to develop automatic approaches, especially at HR, in order to make a maximum of models and results generalizable to the rest of the African continent. The urban expansion model and the population density model will be integrated into the Random Forest algorithm recently developed by DGE-US to disaggregate census population counts from administrative units to 100 m pixels. In addition to classical dissemination tools as participation in dedicated workshop and conferences, publication through web-site and academic peer reviewed journals, the project aims to contribute to the WorldPop project.
Result Expected Scientific Results

Different sets of scientific results are expected.

From a technical point of view, different remote sensing data will be combined in order to develop an automatic approach for urban detection in sub-Saharan Africa. Given that urban areas are more complicated to detect in SSA landscapes compared to developed countries (mainly due to the use of man-made materials for building), we expect to demonstrate how the fusion of optical and SAR data can reach higher levels of accuracy in the African context. We also expect to show that developing a method that is specific to SSA may provide better results than existing global datasets.

At the level of individual cities, our work should contribute to identify the geographical factors that contribute to or limit their extension and to better understand how the spatial structure of sub-Saharan cities is framed by their history and by different environmental and anthropogenic factors.  These, should in turn help us understanding how the structure of the cities is linked to the within-city distribution of human population.

The modelling work made across a wider set of cities should complement these understanding by allowing to assess spatial factors that have a generalizable effect on urban growth across the diversity of situations encountered in SSA. Results should confirm our hypothesis that a limited number of generalizable environmental variables can already predict a large part of urban expansion patterns in SSA.

The project will also help to understand the relationships between HRRS and VHRRS landscape attributes and demonstrate the value of using data of different spatial and temporal resolutions.

Expected Products and Services

The expected products and services (deliverables) include :
  • A robust and rapid method integrating SAR and optical data for the delineation of urban extents in Africa (WP1)
  • A database of land cover change to urban over the last 20 years across 30 cities in Africa (WP1)
  • A robust and advanced method integrating VHRRS stereo optical and VHRRS SAR data for the land cover and land use mapping for African cities (WP2)
  • A database of land cover and land use for 3 cities (WP2)
  • High resolution forecasts of urban expansion across Africa in 2020-2030 (WP3)
  • Intra-urban population density maps for 30 (WP3)
  • Automated procedure to integrate the two models into Africa-wide population datasets (WP3)
  • Map of intra-urban population densities for 3 selected African cities (WP4)
  • Report of the trans-scale analysis between population densities derived from VHRRS and HR data (WP4)
  • Predicted Africa-wide population distribution datasets for 2020 and 2030 that include urban expansion forecasts (WP5)
  • Africa-wide population distribution datasets that better predict intra-urban population densities (WP5)
  • Dissemination of results through WorldPop website, publications, presentations and social media (WP5)
  • A survey that identifies the interested users and their requirements (WP7)

In addition to its own set of scientific objectives, the project aims to contribute to the AfriPop/WorldPop project, which should ensure a rapid dissemination of its deliveries.
Website link
Team Member: PERNEEL Christiaan ERM - KMS - RMA (Ecole Royale Militaire/Koninklijke Militaire School)
Team Member: WOLFF Eléonore ULB - Institut de Gestion de l Environnement et d Aménagement du Territoire (IGEAT)
Project Leader: GILBERT Marius ULB - Lutte biologique et Ecologie spatiale (LUBIES)
Project Leader: LINARD Catherine ULB - Lutte biologique et Ecologie spatiale (LUBIES)
Team Member: SHIMONI Michal ERM - KMS - RMA (Ecole Royale Militaire/Koninklijke Militaire School)
Team Member: TATEM Andrew University of Southampton - Geography and Environment
Sensors used
Applications Urban & suburban
Related presentations BEODay 2016 - 06. MAUPP: Mapping and modelling urban patterns in sub-Saharan Africa through remote sensing and data fusion
Related publications
Datasets used